基于BP神经算法可预测的快速瓦斯传感器研究  

Research of Predictabale and Fast Gas Sensors Based on BP Neural Algorithm

作　　者：梁美玉[1] 郭慧[1] 蒋龙云 龙海燕[1] 

机构地区：[1]安徽工程大学机电学院,安徽芜湖241000

出　　处：《煤矿机械》2016年第3期166-168,共3页Coal Mine Machinery

摘　　要：基于传统的瓦斯传感器检测原理,提出了一种基于BP神经算法的可预测的快速瓦斯传感器。研究以FPGA为核心处理器,在进行当前瓦斯浓度检测的软件设计过程中,加入BP神经算法反演策略,实现可预测的瓦斯传感器功能。传感器可以通过当前时刻的前几个瞬时瓦斯浓度值,预测当前瓦斯浓度的稳态值,实现浓度检测的可预测性,同时也提高了瓦斯传感器的检测速度。Based on the traditional gas sensor detection principle, this paper proposes a predictable and fast gas sensor based on BP neural algorithm. FPGA as the core processor is studied, in the current gas concentration measurement, BP neural algorithm inversion strategy was joined in the process of software design to Achieve the function of predictable.Gas concentration of the steady state value can be forecasted by using the current moment in the first few instantaneous gas density,and the detection speed of gas sensor can be improved.

关 键 词：BP神经算法 FPGA 可预测性 瓦斯传感器 

分 类 号：TD712[矿业工程—矿井通风与安全]